From: Trans <transfire@...>
Date: 2006-08-16T00:15:06+09:00
Subject: Re: nonce


ts wrote:
> >>>>> "T" == Trans  <transfire@gmail.com> writes:
>
> T> Well, I suppose to be exact the inside of the shirt should say
> T> "pro-singleton" and the outside "anti-singleton" --that would be like
> T> Ruby where there is on #singleton method, but under the hood there are
> T> rb_singleton... refrences. But then nobody would see a  "pro-singleton"
> T> message on the inside of a shirt ;)
>
>  This is because you don't want to see it
>
> moulon% ruby -e 'p Kernel.private_methods.grep(/singleton/)'
> ["singleton_method_removed", "singleton_method_undefined", "singleton_method_added"]
> moulon%
>
>  Even ruby know what is a singleton class :-)
>
> moulon% ruby -e 'class << Object; self end.dup'
> -e:1:in `initialize_copy': can't copy singleton class (TypeError)
> 	from -e:1
> moulon%

Yes, you're right. But it has been a slow "turning the shirt inside
out" --differnt names had been used and there is still no direct
#singleton_class method.

And in anycase, I still think it's rather sad that another name can't
be arrived at, so as to avoid terminology clash wtih "that other
singleton". I mean, metaclass was abandoned becuase it was too general
and virtual class because of a clash with those in C. Is it so hard to
pick a name that isn't already being used? And it's not like it's a
word being used in the pottery field or some other totally separate
parlance. Singleton has meaning in this very realm: require
'singleton'.

And besides that:
http://7rans0noma.blogspot.com/2006/07/singleton-2-legion.html

So maybe the whole "singly"' thing isn't so imparative. I've been
thinking about the utiltiy of this "legionary" approach in
prototype-based coding for instance.

T.